Rethinking how to produce video to facilitate rapid iteration led to unexpected benefits.
These days, ‘agile’ is often thought of as a methodology for teams to deliver changes on a continuous basis through a workflow of short phases. Although borne of the software and engineering industries, the theory of Agile is increasingly being applied to other fields and industries. In this post, Ondema’s Co-founder (and resident US Marine veteran) details how the military incorporated this type of thinking to help accelerate situational adaptability.
Ondema, being a Software as a Service (SaaS) company, also leverages agile in developing its primary product, the Workspace. Although this method directly benefits customers, the constant change compounds the workload of video Workspace Tutorial production. The Workspace’s constantly evolving user interface constantly renders these videos out of date.
The traditional video production process is very much a set of serial events. You shoot the video content, edit it, and then publish the project. Even for screen capture type footage that we use in the video tutorials, it has to be captured as a video. Reshooting the video each time there is a small interface change simply isn’t feasible due to the large amount of time required to effectively redo the project. So this begged the question: how can we change the video production process from one of having to repeat large parts (or all) of the video creation process to a modular process where individual components of the video can be quickly updated? The key goal being to reduce the amount of time needed to iterate.
First, we attempted to break up the video production process into smaller video clips enabling us to just reshoot that clip instead of the entire sequence. While it promised to reduce the recording time needed, it suffered several drawbacks:
- Timing: it was difficult to get the new clip to be the same length as the old clip (each mouse move needed to be precisely the same, etc.).
- Continuity: placing the mouse in the same position as the previous clip to make a seamless transition was next to impossible. Furthermore, updates in the software meant slight changes in screen components such as banners being inconsistently sized between clips.
Next, we thought about moving from a screen recording method to an animation method using static screen-shots. This gave us the ability to capture an image of each screen and inserting it into the timeline, animating the ‘mouse’ over the top of these screen shots giving a sense of realism. Updates therefore become a rapid and simple process: just capture the screen that had changed, replacing the old image and re-render. Other benefits arose from this new method including:
- Improved User Experience: we now had the ability to create an oversized mouse pointer and animated ‘clicks’, helping making actions clearer to the audience.
- Timing: with animation, timing can be made perfect so if the new screen change the layout, simply move the mouse destinations to the new location (e.g. if a button to click has moved) within the build software and nothing else before or after is impacted.
The tools we use for this process are:
The new process uses the following tools:
- Google Chrome – resized and used to set the required screen of the Workspace.
- Mac OS-X – to capture the screen image.
- Adobe Photoshop – to clean up any of the captured images (optional).
- Apple Motion – used to create the animation objects. These are created as Apple Final Cut Pro templates (either a generator or a title).
- Final Cut Pro – used (and saved) for the overall edit.
Screenshot below showing Final Cut Pro elements.
This new method has dramatically improved our turn-around for updating the tutorials. What previously took a day or more to update, now can be done in under an hour and the unexpected additional benefits have now become must haves. It’s a reminder to never accept the status-quo.
I wonder where this journey will lead us next?